Julia repository of bundle adjustment problems
Author JuliaSmoothOptimizers
8 Stars
Updated Last
11 Months Ago
Started In
July 2021


Documentation CI Coverage Release DOI
docs-stable docs-dev build-ci codecov release doi

Julia repository of bundle adjustment problems from the Bundle Adjustment in the Large repository

How to Cite

If you use BundleAdjustmentModels.jl in your work, please cite using the format given in CITATION.bib.

Bug reports and discussions

If you think you found a bug, feel free to open an issue. Focused suggestions and requests can also be opened as issues. Before opening a pull request, start an issue or a discussion on the topic, please.

If you want to ask a question not suited for a bug report, feel free to start a discussion here. This forum is for general discussion about this repository and the JuliaSmoothOptimizers, so questions about any of our packages are welcome.


julia> using BundleAdjustmentModels

problems_df() returns a DataFrame of all the problems, their group and other features.

julia> df = problems_df()
74×5 DataFrame
 Row │ name                     group      nequ      nvar     nnzj      
     │ String                   String     Int64     Int64    Int64     
   1 │ problem-16-22106-pre     dubrovnik    167436    66462    2009232
  ⋮  │            ⋮                 ⋮         ⋮         ⋮         ⋮
  74 │ problem-1778-993923-pre  venice     10003892  2997771  120046704
                                                         72 rows omitted

When you get this dataframe you can sort through it to get the problems that you want. For example, if you want to filter problems based on their size you can apply this filter:

julia> filter_df = df[ ( df.nequ .≥ 50000 ) .& ( df.nvar .≤ 34000 ), :]
2×5 DataFrame
 Row │ name                  group    nequ   nvar   nnzj    
     │ String                String   Int64  Int64  Int64   
   1 │ problem-49-7776-pre   ladybug  63686  23769   764232
   2 │ problem-73-11032-pre  ladybug  92244  33753  1106928

You can get the problem name directly from the dataframe:

julia> name = filter_df[1, :name]

fetch_ba_name returns the path to the problem artifact. The artifact will download automatically:

julia> path = fetch_ba_name(name)

You can also get an array of the paths to an entire group of problems

julia> path = fetch_ba_group("ladybug")
30-element Vector{String}:

You can directly construct a nonlinear least-squares model based on NLPModels:

julia> model = BundleAdjustmentModel("problem-49-7776")
BundleAdjustmentModel{Float64, Vector{Float64}}

You can then evaluate the residual and jacobian (or their in-place version) from NLPModels:

julia> using NLPModels
julia> residual(model, model.meta.x0)
63686-element Vector{Float64}:
julia> meta_nls = nls_meta(model)
julia> Fx = similar(model.meta.x0, meta_nls.nequ)
julia> residual!(model, model.meta.x0, Fx)
63686-element Vector{Float64}:

You need to call jac_structure_residual! at least once before calling jac_op_residual!.

julia> meta_nls = nls_meta(model)
julia> rows = Vector{Int}(undef, meta_nls.nnzj)
julia> cols = Vector{Int}(undef, meta_nls.nnzj)
julia> jac_structure_residual!(model, rows, cols)
([1, 163686, 63686], [1, 223768, 23769])

You need to call jac_coord_residual! everytime before calling jac_op_residual!.

julia> vals = similar(model.meta.x0, meta_nls.nnzj)
julia> jac_coord_residual!(model, model.meta.x0, vals)
764232-element Vector{Float64}:
julia> Jv = similar(model.meta.x0, meta_nls.nequ)
julia> Jtv = similar(model.meta.x0, meta_nls.nvar)
julia> Jx = jac_op_residual!(model, rows, cols, vals, Jv, Jtv)
Linear operator
  nrow: 63686
  ncol: 23769
  eltype: Float64
  symmetric: false
  hermitian: false
  nprod:   0
  ntprod:  0
  nctprod: 0

There is no second order information available for problems in this module.

Delete unneeded artifacts and free up disk space with delete_ba_artifact!:

julia> delete_ba_artifact!("problem-49-7776-pre")
[ Info: The artifact ladybug/problem-49-7776-pre.txt.bz2 has been deleted

Use delete_all_ba_artifacts! to delete all artifacts:

julia> delete_all_ba_artifacts!()
[ Info: The artifact dubrovnik/problem-16-22106-pre.txt.bz2 was not found
[ Info: The artifact dubrovnik/problem-88-64298-pre.txt.bz2 was not found
[ Info: The artifact ladybug/problem-138-19878-pre.txt.bz2 has been deleted
[ Info: The artifact ladybug/problem-318-41628-pre.txt.bz2 has been deleted
[ Info: The artifact venice/problem-1408-912229-pre.txt.bz2 was not found
[ Info: The artifact venice/problem-1778-993923-pre.txt.bz2 was not found

Special thanks to Célestine Angla for her initial work on this project during her internship.

Licensed under the MPL-2.0 License

Used By Packages

No packages found.